Measuring predetermined socioeconomic 'inputs' when assessing the efficiency of educational outputs

Data envelopment analysis (a mathematical programming technique) has often been applied to measuring the efficiency with which outputs are produced. The technique derives efficient combinations of outputs for given inputs: constant returns to size may be assumed or one may choose to examine whether decreasing or increasing returns hold true. The technique was originally developed with cost data as the inputs, but many of the applications in education have been required to concern themselves with inputs of a social nature, where the inputs are frequently 'categorized'. This paper develops a method for measuring these categorized inputs in a more satisfactory way. A DEA analysis is undertaken of results for the Local Education Authorities of England in the early 1980s, and is followed by a stochastic frontier analysis.